Neumann and Klotz (1994; Klotz and Wolff 1995) used metacontrast, a form of visual backward masking, to reduce the visibility of the primes to a point where they could not be discriminated. In metacontrast, the visibility of a briefly flashed stimulus is reduced when it is followed by a masking stimulus with adjacent but nonoverlapping contours (Francis 1997; Mack-nik and Livingstone 1998). The critical feature of the task was that prime and mask were similar in shape, so that they shared features relevant for either the correct or the alternative response. The authors found that unidentifiable primes still affected response time (RT) to the masking stimulus, showing facilitatory effects from response-consistent primes and interference effects from inconsistent primes.
We extended these findings to study the time course of the priming effect (Vorberg et al. submitted). We used arrow stimuli as masks that could point either to the left or to the right, and participants made a speeded 2-alternative keypress response to the direction of the mask. Priming stimuli also were arrows, presented in a central cutout of the mask so that their visibility was strongly reduced by metacontrast. Varying the stimulus onset asynchrony (SOA) between prime and mask, we found that a prime pointing in the same direction as the mask speeded responses to the mask, whereas a prime pointing in the opposite direction slowed responses. The priming effect (the response time difference between inconsistent and consistent cases) increased linearly with SOA, yielding a priming function with nearly constant slope. On the other hand, there was no indication of visual awareness of the prime in any condition: Despite extended practice and a highly reliable measurement procedure, participants were not able to discriminate the primes in a forced-choice test.
In another experiment, varying the relative durations of prime and mask presentations yielded qualitatively different time courses of prime identification performance. However, the shape of the priming function was unchanged, regardless of whether the prime could be identified, or whether prime identifiability was an increasing or decreasing function of SOA. These data suggest independence of the time courses of priming and visual awareness, so that the time course of the priming effect depends on the prime-mask SOA only. Since then, these results have proven extremely reliable and have been replicated with different stimuli and types of response.
Schlaghecken and Eimer (1997; Eimer and Schlaghecken 1998) used both event-related potentials and response times to study the physiological dynamics of the priming effect in a task similar to ours. They found that magnitude and direction of the priming effect were paralleled by shifts in lateralized potentials above motor cortex. This strongly suggests that the prime does affect motor responses rather than only visual or semantic aspects of processing.
In short, it seems that the prime directly affects motor performance by activating an assigned response, with the activation level depending only on the time course of visual availability of the prime (the prime-mask SOA). For this reason, I will use the term response priming for this type of priming task with a unique and consistent mapping of primes to responses.
Was this article helpful?